Open lameesnasser opened 4 years ago
Hi, I'm working on a similar issue with small objects. My approach is to segment the image into smaller ones using a sliding window, and then the model is able to detect the objects. You could try that as well.
Hi, I'm working on a similar issue with small objects. My approach is to segment the image into smaller ones using a sliding window, and then the model is able to detect the objects. You could try that as well.
Hi,
I am wondering in the training process. What's the training data size and what's the resolution of your training pics. I also running a small object detection with mask r cnn and can;t get a good result.
Hi, As mentioned in this discussion (https://www.kaggle.com/c/data-science-bowl-2018/discussion/54089) , we can use mask rcnn to segment small objects. However, when I tried mask rcnn on cell adhesion images, it didn't detect the very small objects shown in the figure. Can you please advise me on this issue?